Prostate cancer is a leading cause of cancer morbidity and mortality in the United States. In 1998, it will account for 39,000 deaths with 189,000 new cases being diagnosed. Current screening methodology involves use of digital rectal examination in combination with PSA testing of the blood to identify individuals at risk. This methodology remains unproved at best, and leads to invasive core biopsy in the majority of individuals which is known to have a false negative rate, and which may deter some patients from participating in screening efforts. We propose to initiate studies of two molecular markers known to be of value in detecting prostate cancer cells compared to their normal epithelial counterparts, the expression of telomerase, and the presence of hypermethylated Glutathione Sulfotransferase Pi-1 promoter region DNA. We have established reasonably sensitive methodologies for detection and have shown that they are efficacious in finding LNCaP tumor cells added to normal ejaculates, while not providing a high rate of false positivity. Moreover, we have found that the expression of rare messenger RNA transcripts for the HOXC4-6, C8, and C10 genes characterize prostate cancer cells compared to their normal counterparts. We propose to further evaluate such gene expression in prostate epithelial cells obtained from ejaculates and fine needle aspirates of newly diagnosed patients as a novel diagnostic approach to prostate cancer which could lead to less invasive and/or more informative methods of diagnosis.